A comparative investigation on growth, nanostructure and electrical properties of copper oxide thin films as a function of annealing conditions

This paper is an attempt to compare the influence of various annealing conditions on growth, nanostructure, surface morphology and electrical properties of copper oxide thin films. Cu thin films of 85 nm thickness were deposited on glass substrate by thermal evaporation method, and then post-annealed at different environments (air and oxygen flow), different temperatures (200–400 C) and different times (35 and 75 min). X-ray diffraction results showed (Cu2O) cuprite phase for all annealed samples at 200 and 250 C and (CuO) tenorite phase for all samples annealed under different conditions at 350 and 400 C. A complex phase of CuO and Cu2O was observed for all annealed samples at 300 C, with exception for the sample annealed with flow of oxygen for 75 min. The atomic force microscopy results showed that surface morphology of the samples was strongly affected by the changes of annealing conditions (i.e., time, temperature and environment). The size of the grains increased with annealing temperature and time, while the samples annealed with oxygen flow showed larger grains than those annealed in air. Two different behaviors with annealing temperature were distinguished for the surface roughness of the samples annealed in the air and those annealed with flow of oxygen. Resistivity and Hall effect of samples were measured by a four-point probe instrument and a Hall effect investigation system, respectively. The electrical analyses showed that the variations in annealing conditions had a remarkable effect on measured electrical parameters, namely films resistivity, carriers concentration and type, and Hall mobility.


Introduction
Copper oxide can exist in two crystalline phases viz. cuprous oxide or cuprite (Cu 2 O) and cupric oxide or tenorite (CuO) [1]. Cu 2 O is a p-type semiconductor of cubic structure with a direct band gap of 2.1-2.6 eV, while CuO is a p-type semiconductor having monoclinic structure and an indirect band gap of 1.9-2.1 eV [2]. However, CuO is also reported to possess n-type conductivity [3]. Because of their remarkable properties, both of these materials are widely used in photovoltaic devices [4], solar cells [5], magnetic devices [6] and catalysis [7]. CuO is also used in other various applications, such as gas sensors [5], magnetic storage media [8], solar energy transformation [9], varistors [10].
Many researchers have reported on obtaining copper oxide thin films, using different methods such as sol-gel [11,12], pulsed magnetron sputtering [13], reactive magnetron sputtering [14], reactive evaporation [15], RF sputtering [16], ion beam sputtering [17], molecular beam epitaxy [18], plasma evaporation [19], and CVD [20], and investigation of various properties of these materials. However, a simple method consisting of deposition of Cu thin film and subsequent annealing may be used to produce copper oxide thin films. Figueiredo et al. [3] and Mohemmed Shanid and Abdul Khadar [1] have separately reported on obtaining the copper oxide thin films by this method. They have produced copper oxide thin films by post-annealing of Cu thin films at the air environment and investigated different properties of these layers just as a function of annealing temperature. But, nanostructure and different properties of these films can be affected by other annealing conditions. Therefore, we decided to make use of this method and investigate the influence of different annealing conditions (i.e., annealing temperature, time and environment) on crystallographic structure, surface morphology and electrical properties of the produced samples.

Experimental details
Copper films of 85 nm thickness were deposited on glass substrates (10 9 10 mm 2 cut from microscope slide) by thermal evaporation from tungsten boat. The purity of the copper was 99.998 %. An Edwards (Edwards E19 A3) coating plant was used. Film thickness and deposition rate were measured using a quartz crystal deposition rate controller (Sigma Instruments, SQM-160, USA) positioned close to the substrate. Deposition angle and deposition rate were 158 and 3 Å /s, while base pressure was 2 9 10 -5mbar. Post-annealing of the Cu/glass films was performed by a horizontal tube furnace at five different temperatures of 200, 250, 300, 350 and 400°C in the air environment and oxygen flow of 200 sccm. Two annealing times of 35 and 75 min were used. The samples reached their target temperature with a thermal gradient of 7°/min. Prior to deposition, all substrates were ultrasonically cleaned in heated acetone and then ethanol. Crystallographic information of the produced samples was obtained using a Philips XRD X'pert MPD Diffractometer (Cu Ka radiation) with a step size of 0.02°and count time of 1 s per step. Surface morphology and roughness of the samples were obtained by means of AFM (Park Scientific) analysis with a scan rate of 1 Hz and a scan size of 1 9 1 lm 2 . A four-point probe instrument and a Hall effect system with a magnetic field strength of 0.2 T were employed for resistivity measurements and Hall effect analysis, respectively.

Results and discussion
Crystallographic structure XRD results for all samples are given in Table 1, and the XRD patterns of selected samples (annealed in the air and oxygen flow for 75 min) are shown in Fig. 1 [3] have also reported similar results for their films deposited by electron beam and post-annealed in air for 30 min.
The above-mentioned behavior was observed for almost all samples produced under different conditions, however, with some diversions as follows: Therefore, it may be concluded that the phase transition of copper oxide films produced in this work is directly related to the annealing temperature, environment and time. On the other hand, a higher presence of oxygen and more annealing time promote an oxidation at lower temperatures. For this reason, there was no Cu 2 O phase in the sample which was sintered at 300°C in O 2 environment for 75 min.
Crystallite size (coherently diffracting domains) D was obtained by applying the Scherrer formula (D = kk/ Bcosh), where k is unity, k = 1.54248 Å , h is the peak  [21]. The crystallite size (coherently diffracting domains) results are given in column 7 of Table 1. The results show that grain size increases with annealing temperature, and the films annealed with oxygen flow for longer time have larger grains. The increase in grain size with annealing temperature is clearly due to higher activation energy which causes higher diffusion rate; however, the larger grains obtained for annealing with flow of oxygen relative to those annealed in air can be due to the fact that air apart from oxygen consists of nitrogen and other gases or contaminants that are usually embedded between grains and stop grains from growing larger [22,23].
Surface morphology 2D AFM images of selected samples (i.e., samples annealed at 200, 300, 400°C in the air and oxygen flow for 75 min) are shown in Fig. 2a-f. Grain sizes of all samples were measured from 2D AFM images using JMicrovision Code. Results are given in column 8 of Table 1. These results show that the grain size increases with annealing temperature and when samples annealed with flow of oxygen are compared with those annealed in air, it can be observed that the formers contain larger grains. Figure 3 depicts the 3D AFM images of selected samples, while rms and average roughness of films obtained from AFM analysis are given in columns 9 and 10 of Table 1. These results show that for both annealing environments, namely air and oxygen film, surface roughness increases with annealing time for all annealing temperatures used in this work. This can be due to the fact that at longer annealing time, atoms can move more and fill/ remove the defects of different types (i.e., voids, interstitials and other types), hence larger grains may form as mentioned in column 8 of Table 1. In addition, by formation of larger grains usually the valleys between the grains become wider and deeper (i.e., grooving effect [24,25]), hence surface roughness increases. Moreover, two different behaviors can be distinguished for the samples annealed in the air and those annealed with flow of oxygen. The samples annealed in the air show a minimum surface roughness at 300°C annealing temperature, while those annealed with flow of oxygen show a maximum value at this temperature.

Concentration of carriers and Hall mobility
The Hall's effect is important, because it enables us to make measures of mobility and concentration of carriers, and gives insight into the mechanism of conductivity in semiconductors. Figs. 4 and 5 illustrate the variation of carrier concentration and Hall mobility of our samples, respectively. The numerical data and the type of carriers are also given in Table 2. Results ( Table 2, column 6) show p-type conductivity for samples with Cu 2 O structure, while samples with CuO structure have n-type  Figure 5 and Table 2 (column 7) show that the carrier concentration decreases with increasing the annealing temperature from 200 to 250°C, while it increases with annealing temperature at higher temperatures. This is true for all samples regardless of their annealing environment or the annealing time. The comparison of these results with the phase identification of the samples given in column 5 of Table 1 suggests that this observation may be related to the crystallographic structure changes of the samples with annealing temperature which in turn is related to the carrier type in the samples (i.e., change from Cu 2 O (p-type carrier) to CuO (n-type carrier)).
The variation of Hall mobility for all samples produced in this work is given in column 8 of Table 2 and Fig. 5. From these results and the results given in Table 1, for crystallographic structure of the samples, it can be deduced that these variations are related to the carrier concentration as discussed above, hence are directly influenced by the film crystallographic structure (phase). Consequently, there exists a correlation between the results obtained for carrier concentration, Hall mobility and the crystallographic structure (phase) of the samples produced in this work.

Resistivity
Results of DC resistivity measurement obtained using a four-point probe instrument is discussed in this section. To investigate the influence of possible (low frequency) charging effects at the electrical contacts and leads, current-voltage (I-V) curves were recorded in both increasing and decreasing increments of voltage. The results of the measurements in the decreasing direction of the voltage are not shown as they matched the results in the increasing voltage direction and only clutter the figure unnecessarily. Linear I-V curves were obtained for all samples; independent of the scan direction, there was no indication of  hysteresis effect. To investigate the anisotropy effect in these samples, the I-V curve measurements were also carried out in four different directions on the samples, namely two vertical (along the sample length and normal to the length) and two diagonal directions. Figure 6 depicts the results of I-V curves for selected samples that were heated in air for 35 min as a function of annealing temperature, while the values of average resistivity for all samples measured at room temperature are given in column 5 of Table 2 and Fig. 7. The results show that resistivity increases with annealing temperature up to 300°C and decreases at higher annealing temperatures. The variation of the resistivity of our samples again regardless of their annealing environment and annealing time shows a peak value for the samples annealed at 300°C. Consulting the crystallographic structure (Table 1, column 5) and carrier type ( Table 2, column 6) indicates that at this temperature the samples contain a mixture of two phases with two different types of carriers, namely Cu 2 O (p-type) and CuO (n-type), though the final identification in Table 2 for the carrier type of these samples is given as p-type. The increased resistivity of these samples could be due to existence of these two different types of carriers in these samples that act in two different directions.
A similar variation of resistivity for films produced by e-beam deposition and subsequent annealing in air for 30 min is reported by Figueiredo et al. [3].  ) for all samples, respectively, while at the transition annealing temperature of 300°C a mixture of both of these phases was obtained for all samples except that annealed with flow of oxygen and for 75 min. Grain sizes obtained from AFM images indicated that the grain size increased with annealing temperature and annealing time, while those annealed with flow of oxygen produced larger grains than those processed in air. Samples annealed in the air showed a minimum surface roughness at 300°C annealing temperature, while those annealed with flow of oxygen showed a maximum surface roughness at this annealing temperature. This phenomenon may be related to the crystallographic structure of these films, so that those annealed at this certain temperature and in air contain both CuO and Cu 2 O grains, whereas those annealed in the oxygen flow (at 300°C) contain only CuO grains (particularly at increased annealing time of 75 min) or much less amount of Cu 2 O (35 min annealing time). This can also be deduced from the ratio of intensities of the peaks of Cu 2 O to CuO for these samples (Table 1, column  7), hence distribution of smaller grains (lower intensities of three peaks in the XRD pattern) causes almost a uniform surface (minimum surface roughness) while large grains (higher intensities of two peaks in the XRD pattern) produce a rough surface. In fact, our discussion emphasis is on the distribution of smaller Cu 2 O grains among the CuO grains. The electrical results showed n-type conductivity for CuO phase and p-type conductivity for Cu 2 O phase. The variations in the resistivity, concentration of carriers and Hall mobility showed a direct correlation with the crystallographic phase changes of the samples as a function of annealing temperature regardless of annealing time and annealing environment.